AI study finds new patient reported drug side effects.
A new AI study has found that artificial intelligence can help uncover possible GLP-1 drug side effects, popular weight loss and diabetes drugs by analyzing what people share online. Researchers reviewed more than 400,000 posts written by nearly 70,000 users on Reddit over several years. The focus was on medications known as GLP-1 drugs, including semaglutide and tirzepatide, which have become widely used in recent years.
These medications are commonly prescribed to help manage blood sugar and support weight loss. While clinical trials have already identified many known side effects, such as nausea and stomach problems, researchers wanted to see if patient experiences shared online might reveal additional concerns. Social media platforms often include personal accounts that may not always be reported during doctor visits or captured in formal studies.
Using advanced computer tools, the research team sorted through large amounts of text to identify patterns in how people described their symptoms. This approach allowed them to connect everyday language used in posts with standard medical terms. The method made it possible to review a much larger amount of data than traditional approaches, which often require manual review and are limited in scale.
The findings showed that many users reported expected side effects, especially digestive issues. About 44 percent of users mentioned at least one side effect, which helped confirm that the method was detecting real and known reactions. However, the AI study also covered GLP-1 drug side effects that may not be fully reflected in current drug information.
Among these were reports related to reproductive health. Some users described changes in menstrual cycles, including irregular timing, heavier bleeding, or spotting between periods. While only a small portion of users reported these symptoms, the number was large enough to raise interest among researchers. These reports suggest a need for further study to understand whether there is a link and how often it may occur.
Another group of symptoms involved changes in body temperature. Some users said they experienced chills, feeling unusually cold, hot flashes, or fever-like sensations. These types of complaints are not widely discussed in official drug information but appeared often enough in the online data to stand out.
Fatigue was also commonly reported and ranked as one of the top concerns among users. Even though tiredness has appeared in some clinical findings, it has not always been a major focus. The frequency of these reports online suggests that it may have a bigger impact on daily life than previously understood.
Researchers emphasized that the AI study does not prove that these drugs directly cause the newly identified symptoms. Social media data cannot confirm cause and effect, and the group of users studied may not represent the general population. Reddit users tend to be younger and are more likely to be based in the United States, which may influence the types of experiences shared.
Despite these limits, the AI study shows the value of looking at patient conversations online. People often share details in these spaces that might not come up during medical appointments. These discussions can act as early signals, helping researchers notice patterns that may deserve closer attention in future studies.
The use of artificial intelligence made it possible to carry out this kind of large-scale review more quickly than in the past. As more medications gain attention and are used by larger groups of people, this type of monitoring may help identify concerns sooner. It may also guide researchers toward areas that need more formal investigation.
The team plans to expand the research to include other platforms and different groups of users. This may help determine whether the same patterns appear across wider populations. As interest in GLP-1 drugs continues to grow, ongoing tracking of patient experiences may play an important role in understanding their full range of effects.
Sources:
AI analyzes Reddit posts to find underreported GLP-1 side effects
Self-reported side effects of semaglutide and tirzepatide in online communities
Join the conversation!